Cheng (1986) trained male rats to search for food in a rectangular arena that also contained distinctive visual patterns. He found that the rats used mainly the geometric framework of the box itself to find the food and claimed that geometrical information is processed in a specialized module, which is independent of feature information. The aim of the present set of experiments was to check if the previous results with male rats and an appetitive task could be extended to an aversive task while using both male and female rats and three-dimensional landmarks. In Experiments 1 and 2, rats were trained in a rectangular-shaped pool to find a hidden platform that had a location defined in terms of two sources of information—landmarks of different salience (less salient in Experiment 1, more salient in Experiment 2) outside the pool and a particular corner of the pool. The results showed that both males and females used mainly the particular corner of the pool, supporting Cheng’s suggestion. In addition, in the two experiments, what the rats learned with respect to the landmarks was negligible. Experiment 3 used a more difficult triangular pool in addition to the rectangular pool, in the absence of landmarks. The results revealed sex differences in the triangular pool but not in the easier rectangular pool. These results suggest that task difficulty is a factor when it comes to finding sex differences in rats in spatial tasks.